The long-range goal of this research is to improve understanding of small-scale mixing processes in the atmospheric boundary layer and to incorporate the effects of these processes in mesoscale models. Studies of the atmospheric boundary layer using large-eddy simulation (LES) have demonstrated the value of these models in describing basic turbulent processes in the atmospheric boundary layer. We are now at a point where LES can be applied to a broader range of problems that include decoupled flow in stable boundary layers and cases with strong baroclinic shear coupled with convection. This proposal describes a study to examine the role of turbulent mixing in defining boundary layer structure during conditions with weak winds and stratification and cases with strong vertical momentum flux. LES experiments will be used in combination with measurements taken during the Coupled Boundary Layers Air Sea (CBLAST) field programs to test existing non-local mixing schemes and examine alternatives when these schemes fail. Boundary layer parameterizations developed in this study will provide the connection between surface flux algorithms developed as part of CBLAST, with operational mesoscale models such as the COAMPS forecast model. Our goal is to increase the accuracy of coastal mesoscale prediction by adding physically-based approximations to onedimensional mixing parameterizations.